Pancreatic ductal adenocarcinoma is among the most lethal malignancies. Early detection is vital for patients to receive potentially curative surgery, but pancreatic cancer is often asymptomatic at the earliest most curable stages. Thus, the identification of new predictive tumor biomarkers is essential to permit the detection of early-stage pancreatic cancer. We will identify novel biomarkers of pancreatic carcinoma using a two stage approach. First, we will identify proteomic changes in the circulation of a spontaneous mouse model that replicates human pancreatic cancer genetics and disease. Mice, genetically predisposed to pancreatic cancer, will be metabolically labeled in vivo to allow quantitative analysis of the plasma proteome. Serial plasma samples collected before and during tumor development will be analyzed by Multidimensional Protein Identification Technology (MudPIT) to identify proteins that are elevated during the onset and/or progression of disease. MudPIT is a quantitative and highly sensitive proteomics platform that facilitates the discovery of differentially expressed proteins in complex samples. Biostatistical analysis of the MudPIT data will identify markers that display a significant change in circulating levels during disease progression in the mouse. To identify predictive markers of human disease, plasma from patients will be assayed by multiple reaction monitoring (MRM) for the presence of the statistically relevant mouse markers. MRM is a mass spectrometry (MS)-based approach that allows the simultaneous quantitation of multiple target analytes in samples such as plasma. In a second approach we will directly identify pathology-associated biomarkers in murine and human tumor biopsies using a novel activity-based proteomics method. This approach will identify enzymatic activities that show statistically meaningful changes as a function of tumor progression. We will validate the physiological role of enzymatic activities that change concomitant with disease progression through genetic modulation in a fluorescence orthotopic mouse model of pancreatic cancer. Identification of novel tumor-associated biomarkers may enable detection of pancreatic cancer at the earliest stages and facilitate the development of novel therapeutic strategies for this lethal disease. Pancreatic cancer often develops without obvious symptoms and since an accurate & specific screening test does not exist, diagnosis is frequently made too late for surgery to be an option. Using two state-of-the-art analytical methods, we will examine blood samples and tumor biopsies to identify markers that indicate the presence of pancreatic cancer at the earliest stages. These studies could result in a blood-based screening test for early detection of pancreatic cancer, as well as novel therapeutic strategies that could profoundly influence disease outcome. [unreadable] [unreadable] [unreadable]